Poetry

提供诗词生成与诗词查询两大核心能力。 基于 chinese-poetry 开源数据集（收录唐诗 57,607 首、宋诗 254,248 首、宋词 21,053 首、 元曲 11,057 首、诗经 305 篇、楚辞 65 篇、论语 20 章，以及四书五经、蒙学经典、纳兰性德诗集、 曹操诗集等），为用户提供精准的诗词...

Security Scan

This skill is internally consistent: it uses a local chinese-poetry dataset downloaded via a provided setup script and its requirements/instructions match its stated purpose of generating and querying classical Chinese poetry.

Purpose & Capability

The name and description claim poetry generation and dataset-backed querying; the skill requires the chinese-poetry dataset and the SKILL.md and PROMPT.md describe using data/ for lookups and for generation guidance — these needs align with the stated purpose.

Instruction Scope

Runtime instructions are prompt-based and limited to either generating poems (using the provided prompts/formatting rules) or querying the local dataset. The only external action instructed is running setup.sh to download the dataset; there are no instructions to read unrelated system files, environment secrets, or to transmit data to unexpected endpoints.

Install Mechanism

There is no package install spec; the included setup.sh performs a git clone from the public GitHub repository (trusted source). This is a reasonable approach for a large dataset. One minor incoherence: the skill's metadata lists no required binaries, but setup.sh requires git and network access to run — the README and SKILL.md mention running the script but do not declare git as a prerequisite.

Credentials

The skill requests no environment variables, credentials, or config paths. The prompt and docs do not reference secrets or unrelated credentials; requested access is proportional to the described functionality.

Persistence & Privilege

always is false and the skill is user-invocable; it does not request permanent agent inclusion or modify other skills or system-wide settings. The only filesystem change is the dataset cloned into data/, which the user explicitly initializes via setup.sh.

Assessment

This skill appears coherent and matches its description, but before installing: (1) ensure git and network access are available because setup.sh clones the dataset (the skill didn't declare git as a required binary); (2) be aware the dataset download is ~291MB and will be stored under data/; (3) review the chinese-poetry license if you have legal/redistribution concerns; and (4) if you are concerned about autonomous agent behavior, remember the skill can be invoked by the agent (default) — combine that with your agent policies if needed.

Poetry Skill — 中华古典诗词

概述

本 Skill 是一个 Prompt-based 自然语言技能，专注于中华古典诗词领域，具备两大核心能力：

1. 诗词生成 — 根据用户指定的主题、风格、体裁等要求，生成符合古典诗词格律规范的原创作品
2. 诗词查询 — 基于本地完整数据集，按作者、标题、关键词、朝代等条件检索古典诗词

数据来源

本 Skill 依赖 chinese-poetry 开源数据集，数据存储于 data/ 目录。

⚠️ 首次使用前需初始化数据：data/ 目录未包含在发布包中（体积约 291MB），请运行以下命令下载：

./setup.sh

使用方法

诗词生成

通过自然语言描述你的需求即可触发诗词生成：

请帮我写一首关于春天的五言绝句
以"月下独酌"为题，写一首七言律诗
请以豪放派风格填一首《水调歌头》
用婉约风格写一首关于离别的词

诗词查询

通过自然语言描述你想查找的诗词信息：

查一下李白的《静夜思》
帮我找苏轼的所有词作
搜索包含"明月"的唐诗
有哪些关于"秋天"的宋词？
诗经中有哪些描写爱情的篇章？

进阶用法

对比李白和杜甫在"月"这一意象上的不同表达
分析《离骚》的修辞手法
列出辛弃疾最经典的豪放词十首
帮我从纳兰性德的词中找出适合表达思念的句子

文件结构

poetry/
├── SKILL.md          # Skill 元信息与使用说明（本文件）
├── README.md         # 项目说明文档
├── PROMPT.md         # 核心 Prompt 指令（诗词生成与查询规范）
├── setup.sh          # 数据下载初始化脚本
├── .gitignore        # 忽略 data/ 目录
├── LICENSE           # 开源许可证
└── data/             # 诗词数据集（需通过 setup.sh 下载）
    ├── 全唐诗/        # 唐诗 57,607 首 + 宋诗 254,248 首 + 作者信息
    ├── 宋词/          # 宋词 21,053 首 + 词人信息
    ├── 诗经/          # 诗经 305 篇
    ├── 楚辞/          # 楚辞 65 篇
    ├── 论语/          # 论语 20 章
    ├── 元曲/          # 元曲 11,057 首
    ├── 四书五经/      # 大学、中庸、孟子
    ├── 蒙学/          # 三字经、千字文、百家姓、唐诗三百首等
    ├── 纳兰性德/      # 纳兰性德诗集 258 首
    ├── 曹操诗集/      # 曹操诗集
    ├── 五代诗词/      # 花间集、南唐词
    ├── 幽梦影/        # 幽梦影
    ├── 水墨唐诗/      # 水墨唐诗
    ├── 御定全唐詩/    # 御定全唐詩
    └── strains/       # 诗词平仄格律数据